Breaking: Cetuximab inferior to Cisplatin when combined with XRT for HPV+ Oropharynx CA

[Gfunk6]

Surprising results from NRG


https://www.nrgoncology.org/Portals...splatin - FINAL.pdf?ver=2018-08-14-095719-720


Sent from my iPhone using SDN mobile

---

Members don't see this ad.

 

[evilbooyaa]

Not at all surprised, IMO. Cetuximab to me was always the "patient can't tolerate full dose cisplatin" chemotherapy option for OPhx.

Hopefully med oncs will stop doing Cetuximab for everybody and really only do it for who they should: patients who can't tolerate bolus cisplatin. Now the question is going to be from the med oncs whether Cetux or weekly Cis is better given that weekly Cis (at 30mg/m2) is inferior to bolus Cis in locally advanced HNSCC per the Noronha JCO trial.

I personally think Cetux is less efficacious than even weekly Cis (when it's dosed correctly at 40mg/m2), but I don't have a randomized trial to tell me that.

 

[medgator]

Glad the preferred mantra was for cisplatin in these patients until proven otherwise. Thanks for the update

 

[medgator]

.

Hopefully med oncs will stop doing Cetuximab for everybody and really only do it for who they should: patients who can't tolerate bolus cisplatin. Now the question is going to be from the med oncs whether Cetux or weekly Cis is better given that weekly Cis (at 30mg/m2) is inferior to bolus Cis in locally advanced HNSCC per the Noronha JCO trial.

I personally think Cetux is less efficacious than even weekly Cis (when it's dosed correctly at 40mg/m2), but I don't have a randomized trial to tell me that.

Agree. My med oncs will usually run concurrent therapy options by me before we get started.

Unless there are kidney issues, we avoid erbitux. It's not a benign drug either with the skin and mucosal toxicity we see with it. Nearly every patient ends up getting doxy from the MO because of those skin sx.

 

[ramsesthenice]

It's not a benign drug either with the skin and mucosal toxicity we see with it.

This. I like Dr. Bonner but I honestly don't know how they got the toxicity data that they did. There are multiple, well-documented series of greatly increased grade IV mucosal toxicity with concurrent cetuximab. I think that it is a good alternative for patients who absolutely can't get cisplatin for very specific reasons BUT I think it is absolutely wrong to interpret it as a "gentler" alternative to cisplatin because it is definitely not. My current and former institutions have largely abandoned concurrent cetuximab except as a matter of last resort for head and neck SCCs.

The single worst case I have ever had was a 60 something VIP of the hospital. She had a T2N1 OP SCC and I wanted to do RT only but one of her associates who was a higher-ranking physician than me convinced her to do concurrent drug "just in case." She had tinitus so she got concurrent cetixumab. By mid 40s she had severe mucosal toxicity, aspirated, and died.

 

[evilbooyaa]

Hmm, haven't seen that much in terms of mucosal toxicity with cetux (or anything more than just standard treatment). Interesting. My main concern was always oncologic efficacy given that cetux has failed to replace chemo in every other site it was tried in.

 

[RadOncMegatron]

This should make us all pause really fast on de-escalation.

 

[thaddeus]

Wow that is surprising, especially since the trial used the DHANCA regimen, which many people thought would be sufficient for HPV-associated HNSCC without any systemic therapy (See HN002). Hope my med oncs see this....

 

[Haybrant]

This should make us all pause really fast on de-escalation.

Explain further

 

[medgator]

Explain further

Erbitux is thought to be less toxic, but less efficacious in h&n scc. This trial just showed that erbitux really is inferior, even in a population of better acting hpv scc where a less effective treatment should have been enough.

AFAIK, many of the de-escalation protocols have been smaller and single arm studies with shorter fu. Why screw around with something that works really well (seriously, I think I can count my h&n/OP scc failures on one hand since training, and I follow my patients for 5 years) until we know for sure that such endeavors are safe?

 

[Ray D. Ayshun]

Erbitux is thought to be less toxic, but less efficacious in h&n scc. This trial just showed that erbitux really is inferior, even in a population of better acting hpv scc where a less effective treatment should have been enough.

AFAIK, many of the de-escalation protocols have been smaller and single arm studies with shorter fu. Why screw around with something that works really well (seriously, I think I can count my h&n/OP scc failures on one hand since training, and I follow my patients for 5 years) until we know for sure that such endeavors are safe?

Perhaps this is rhetorical, but are HPV-driven cancer less dependent on EGFR signaling than ole-fashioned H&N cancers?

 

[Palex80]

Well, this is HPV positive disease which has its own behaviour, so results should be not generalized in terms of what happens with HPV negative disease.

I'd be interested in looking into patterns of recurrence.

Interestingly, if you look at the original Bonner paper on Cetuximab, you will see that the effect Cetuximab had all came out of better control at the primary tumor site and the regional nodes. It was not distant metastasis, in fact distant metastasis was more or less the same with or without cetuximab. Now, we have some evidence pointing out that HPV-positive tumors are more prone to cause distant metastasis as first site of failure. Why? Because local control is quite good with radiotherapy or radiochemotherapy. Patients with HPV-negative recurrence often do not develop metastasis until very late into the disease course. They show tumour progression locally and some die because of it before developing mets.

Perhaps this is what happened in the NRG trial too. Radiotherapy (standard dose of 70 Gy with slight acceleration in 6 weeks, some may call this an overkill for HPV-positive disease...) perhaps controlled the neck quite good with cisplatin or cetuximab, but the patients recurred at distant / non-irradiated sites? It's only speculation on my part but perhaps something to consider.

Another issue is the choice of fractionation. 70 Gy in 6 weeks is fine, but as some of you may recall, in non-planned subgroup analysis of the Bonner trial Cetuximab only worked in the patients getting concomitant boost. In the other fractionation schedules the effect of cetuximab was considerably smaller. Perhaps cetuximab "needs" hyperfractionation in order to work? Pure speculation again...

 

[RadOncMegatron]

Explain further

First, I must give my mea culpa here as one of the ones who drank the de-escalation Kool Aid...

We, as a field, were so high on so many attempts at de-escalation - lowering XRT dose, selecting by induction, TORS only arms, lowering chemo, lowering treatment volumes in multiple phase II trials.

Now alas, a phase III trial has come where full dose chemorads vs full dose cetuximab where an arm with accelerated XRT with a systemic agent fell flat. There were even talks of accelerated XRT alone for some of these patients based on MACH-NC, but now after years of hearing about "wait for phase III data" it turns out that the good ol' standard of care is better.

I am really saddened by this data and eagerly await the completed manuscript. I really still want to believe in de-escalation (silly me?)!

 

[nkmiami]

A number of years ago at ASTRO, one of the physicists from Wisconsin had a good presentation on just how bad phase II data and retrospective series are at predicting subsequent randomize trials.

 

[temujim]

Easy bro. 1016 was all p16+ oropharynx comers (except for T1-2 N0-1 -- for goodness sake don't give chemo to those folks). Let's see what the subgroups show... the 0129 low risk vs intermediate risk or AJCC 8 stage I vs II vs III or even per protocol vs not per protocol. And dare I say high volume centers vs low volume? Anyway, huge trial. There will be many ways to slice this data I'm sure.

First, I must give my mea culpa here as one of the ones who drank the de-escalation Kool Aid...

We, as a field, were so high on so many attempts at de-escalation - lowering XRT dose, selecting by induction, TORS only arms, lowering chemo, lowering treatment volumes in multiple phase II trials.

Now alas, a phase III trial has come where full dose chemorads vs full dose cetuximab where an arm with accelerated XRT with a systemic agent fell flat. There were even talks of accelerated XRT alone for some of these patients based on MACH-NC, but now after years of hearing about "wait for phase III data" it turns out that the good ol' standard of care is better.

I am really saddened by this data and eagerly await the completed manuscript. I really still want to believe in de-escalation (silly me?)!

 

[medgator]

Easy bro. 1016 was all p16+ oropharynx comers (except for T1-2 N0-1 -- for goodness sake don't give chemo to those folks). Let's see what the subgroups show... the 0129 low risk vs intermediate risk or AJCC 8 stage I vs II vs III or even per protocol vs not per protocol. And dare I say high volume centers vs low volume? Anyway, huge trial. There will be many ways to slice this data I'm sure.

I have no qualms giving my p16+ smokers with T2N1 disease chemorads, personally

 

[Burt Radnolds]

Easy bro. 1016 was all p16+ oropharynx comers (except for T1-2 N0-1 -- for goodness sake don't give chemo to those folks). Let's see what the subgroups show... the 0129 low risk vs intermediate risk or AJCC 8 stage I vs II vs III or even per protocol vs not per protocol. And dare I say high volume centers vs low volume? Anyway, huge trial. There will be many ways to slice this data I'm sure.

I personally recall enrolling many patients on this trial while in residency. I remember seeing a few patients that had N3 disease engulfing half their head and neck randomized to cetuximab and cringing. Patients such as these may have driven the overall outcomes a bit, will definitely be very interested in the actual manuscript.

 

[nkmiami]

Anecdotally, pts seem to handle 100 cis better today than 15 years ago.

 

[medgator]

Anecdotally, pts seem to handle 100 cis better today than 15 years ago.

Age dependent imo, I've treated 30-50 year olds that crash and burn, while the 65+ crowd seems to tolerate things better

 

[BobbyHeenan]

Age dependent imo, I've treated 30-50 year olds that crash and burn, while the 65+ crowd seems to tolerate things better

Agree.

Something about that combo of young, full dose cis, and thick secretions that cause gagging has had me managing patients with intractable vomiting requiring frequent IVFs or hospitalizations.

Then I'll have some 60 year old VA guy fly through it like it's nothing.

Hard to predict.

Definitely think the weekly cis do better, but their tumors probably do better (ie fail) too.

 

[Palex80]

Definitely think the weekly cis do better, but their tumors probably do better (ie fail) too.

Level I evidence for that.
Once-a-Week Versus Once-Every-3-Weeks Cisplatin Chemoradiation for Locally Advanced Head and Neck Cancer: A Phase III Randomized Noninferiority Trial. - PubMed - NCBI

One could argue that 30 mg/week was too low in the trial and thus the cumulative dose of cisplatin lower in the weekly-arm than in the 3-weekly-arm. On the other hand, if you look into the data, you will see that a large proportion of the patients in 3-weekly-regime did not receive all 3 cycles. This is often what many of us (i guess) see in daily practice too: many patients are unfit to receive the 3rd cycle of 3-weekly cisplatin, some only receive only two cycles, some switch to carboplatin, etc...

 

[evilbooyaa]

Level I evidence for that.
Once-a-Week Versus Once-Every-3-Weeks Cisplatin Chemoradiation for Locally Advanced Head and Neck Cancer: A Phase III Randomized Noninferiority Trial. - PubMed - NCBI

One could argue that 30 mg/week was too low in the trial and thus the cumulative dose of cisplatin lower in the weekly-arm than in the 3-weekly-arm. On the other hand, if you look into the data, you will see that a large proportion of the patients in 3-weekly-regime did not receive all 3 cycles. This is often what many of us (i guess) see in daily practice too: many patients are unfit to receive the 3rd cycle of 3-weekly cisplatin, some only receive only two cycles, some switch to carboplatin, etc...

Yeah, that's the Noronha JCO paper I was referencing in my previous post. The main argument from believers of weekly cis is that 30mg/m2 is too low, as 30 x 7 = 210, whereas 100 x 3 = 300 - the argument is that 40 x 7 = 280 and would be much closer.

 

[medgator]

Yeah, that's the Noronha JCO paper I was referencing in my previous post. The main argument from believers of weekly cis is that 30mg/m2 is too low, as 30 x 7 = 210, whereas 100 x 3 = 300 - the argument is that 40 x 7 = 280 and would be much closer.

Yeah 40 weekly is what is used in cervix

 

[evilbooyaa]

Yeah 40 weekly is what is used in cervix

Although that's the standard in cervix (and not bolus like in H&N), there's a similar argument that bolus cis dosing works better in cervix too.

 

[Phantom1]

Yeah, that's the Noronha JCO paper I was referencing in my previous post. The main argument from believers of weekly cis is that 30mg/m2 is too low, as 30 x 7 = 210, whereas 100 x 3 = 300 - the argument is that 40 x 7 = 280 and would be much closer.

Why do people think it is just a simple multiplication? Biological phenomena are usually nonlinear. Of course we know that 200cGyx25 = 50Gy is not the same as 10Gy x 5 = 50Gy. So why do we think that for chemo it is just a simple addition or multiplication??

Look at what the data is showing. Concurrent chemoradiation with bolus cisplatin 100mg/m2 has held up as the standard of care and resulted in superior OS in 2 recent trials in head and neck cancer...

 

[nkmiami]

Yeah, that's the Noronha JCO paper I was referencing in my previous post. The main argument from believers of weekly cis is that 30mg/m2 is too low, as 30 x 7 = 210, whereas 100 x 3 = 300 - the argument is that 40 x 7 = 280 and would be much closer.

What I found strange about this trial was the magnitude of the local failure difference between the 2 arms is almost greater than that seen with radiation alone +/- chemo, so is weekly chemo has almost zero activity?

 

[Palex80]

What I found strange about this trial was the magnitude of the local failure difference between the 2 arms is almost greater than that seen with radiation alone +/- chemo, so is weekly chemo has almost zero activity?

Well, there was no difference in OS between the 2 groups. If weekly chemo had zero efficacy, you would expect the curves to separate by at least 5%. Cisplatin 3 x 100 mg probably has 8% absolute OS benefit, according to the metaanalysis.

 

[evilbooyaa]

Why do people think it is just a simple multiplication? Biological phenomena are usually nonlinear. Of course we know that 200cGyx25 = 50Gy is not the same as 10Gy x 5 = 50Gy. So why do we think that for chemo it is just a simple addition or multiplication??

Look at what the data is showing. Concurrent chemoradiation with bolus cisplatin 100mg/m2 has held up as the standard of care and resulted in superior OS in 2 recent trials in head and neck cancer...

I'm not saying that I agree with the weekly cisers. I'm just stating what they are holding on to in regards to the Noronha trial. I agree with you in principle, but I don't think chemo for radiosensitization works in the same manner as primary radiation. In fact, I'd even consider the argument that more frequent radiosensitization should work better than therapeutic doses given concurrently with radiation, but only once every 3 weeks.

 

[kimplera]

Perhaps this is rhetorical, but are HPV-driven cancer less dependent on EGFR signaling than ole-fashioned H&N cancers?

This MAY be part of the issue - HPV+ cancers have higher rates of activating mutations in the PI3K pathway which may, similar to RAS mutations in colon cancer, be counterproductive with EGFR targeted therapies.

Our data suggests that HPV+ cancers can be radiosensitized by cetuximab in addition to cisplatin. Thus I expected these to be equivalent in the trial. If 70 Gy + cetuximab is undertreatment then we should be really careful with de-escalation. The data will shed light on this - was it only one subset of patients who cetuximab was much worse? If not, this will need to be a major warning flag for the field.

 

[seper]

We should be really careful with any de-escalation.

 

[thecarbonionangle]

HPV+ cancers may not be as sensitive to cetuximab. In cervical cancer for example, even though all trials were done on recurrent or persistent disease, the cetuximab trials were negative. I think if you really want to "de-escalate" the answer may be in combining XRT with different systemic therapies. Why is it that anal cancer requires far less dose? could this also be the game changer in cervical cancer as well? I'm personally not impressed by cetuximab at all.

 

[OTN]

Does dose de-escalation really need to be this difficult? Instead of 70 Gy with concurrent cis, why not 60 Gy? 64 Gy? I don't understand why they're not taking the simplest route to de-escalation, rather than offering up induction chemo (crap) and erbitux.

I'm not going to change a thing with respect to HPV-positive H+N ca tx until we get good data suggesting it's safe to do so. Until then, why move away from a tx paradigm that works so well?

 

[Radiator20]

While I'm a firm believer in bolus cis when possible based on the data we have so far, I've never understood *why* bolus dosing is better. If the goal is radiosensitization, bolus dosing seems like the worst possible schedule, whereas weekly would seem better. Maybe the cisplatin DNA adducts are quite durable (lasting at least 3 weeks between cycles)? Or maybe there's a threshold effect that the bolus 100 gets you over that weekly 40 doesn't? Open to suggestions...

 

[nkmiami]

Does dose de-escalation really need to be this difficult? Instead of 70 Gy with concurrent cis, why not 60 Gy? 64 Gy? I don't understand why they're not taking the simplest route to de-escalation, rather than offering up induction chemo (crap) and erbitux.

I'm not going to change a thing with respect to HPV-positive H+N ca tx until we get good data suggesting it's safe to do so. Until then, why move away from a tx paradigm that works so well?

I am not sure either. It is a real failure of thought leaders and coop groups. Instead, we have induction strategies, weekly chemos, but not straight up dose de-escalation with q3 weekly cis.

 

[medgator]

I am not sure either. It is a real failure of thought leaders and coop groups. Instead, we have induction strategies, weekly chemos, but not straight up dose escalation with q3 weekly cis.

Kinda like how the cooperative group decided to study 60 vs 74 gy in stage 3 lung while many of us were treating 63-66.